Alley in Pisa, Italy

by Lars Kotthoff

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

I was reading about crop factor from various places on the internet, and the general idea that kept building in my mind was that full-frame cameras are good for wide-angle shots but not that good for tele-photo. In Wikipedia it's written:

[...] a 200 mm lens on a camera with a crop factor of 1.5 has the same angle of view as a 300 mm lens on a full-frame camera. The extra "reach", for a given number of pixels, can be helpful in specific areas of photography such as wildlife or sports.

It makes sense, but how high would the level of detail be on a full-frame with 300mm compared to the other example?

share|improve this question
2  
Without at least some idea of specific cameras and lenses to compare, I don't think this can be answered. A 12.8 MP FF camera with a crappy 3rd party 300mm lens will have a lot less detail than a 24.6 MP FF camera with a really great 300mm lens. –  Jerry Coffin Jun 3 '11 at 6:35
1  
@Jerry Coffin - I would assume we are talking about either the same lens, or as close to the same lens as possible for this question. They aren't asking about lenses here really, just the sensor. –  dpollitt Jun 3 '11 at 13:32
    
@dpollitt: Even if we assume that, it doesn't help a lot. A lens with good center but poor edge resolution might favor a crop camera. A lens with lower but more uniform resolution might favor a full-frame. –  Jerry Coffin Jun 3 '11 at 14:52

8 Answers 8

up vote 15 down vote accepted

One aspect of this comparison that has not been mentioned is the fact that crop-sensor cameras are generally faster for a given price point.

  • The 7D shoots at up to 8 Frames/Sec, the 5D manages 3.9 Frames/Sec
  • The 1Ds III manages 5 Frames/Sec, while the 1D III/IV manages 10 Frames/Sec

In sports photography, where continuous drive is often used, those extra frames could mean the difference between capturing a player right before hie hits/kicks a ball, and actually at the moment of contact.


Regarding using telephoto lenses on a crop-sensor body, the critical measurement here is the pixel pitch, which is the thing that actually determines how much detail you will get from a lens.

Basically, if you have two different sensors with the same pixel pitch, the larger one effectively takes the exact same image, with some cropping.

As an example, I have a 30D and a 5D2. Both have 6.4µm pixels. Therefore, every exposure on the 5D2 effectively includes the entire area that a 30D exposure would capture, with the same resolution.

However, the 7D has 4.3µm pixels, so for a given focal length, the 7D will resolve 1.5X (1.488 to be exact) the detail.

This is all assuming an ideal lens. If your lens cannot resolve fine details, either camera will produce a blurry result. Also, small pixel sizes will be less forgiving of lens defects than larger pixels, since the smaller pixels require a larger lpm from the lens. A lens that is at the edge of it's resolving capabilities on a 5D2 may not see any improvement on a 7D, since the extra pixel resolution has no effect on the lens sharpness.

There is a nice breakdown of the Canon series camera pixel pitches on the-digital-picture.com. It's about 2/3 of the way down the page.


There are other considerations - larger pixels generally give less ISO noise, though modern image processing is advancing faster then sensors are shrinking, so it is not as much of an issue as it could be.


Note: I am writing about canon bodies because I am a canon user, and know them far better. However, most of the arguments are much more broadly applicable, basically to anything that uses a CCD/CMOS image sensor:

  • The two critical factors in a camera's FPS are Number of pixels, and ADC speed.
  • The critical factor in how much detail you get from a lens is pixel pitch (Smaller pixels gets more detail, until you reach the limitations of the lens).
  • The largest influence in ISO performance is pixel size (larger pixels are less noisy). The readout electronics have much more influence here, though, so it is not an absolute determining factor as the two above are.

This is true across all brands.

share|improve this answer
    
I don't think comparing the 7D and the 5D Mark II is fair. The 5D is not intended for sports - it is generally slow (including the AF system). Instead, you could look at the D700: It is similarly priced to the 5D and, with a battery grip, equally fast as the 7D. –  eWolf Jun 3 '11 at 8:49
5  
@eWolf: The problem with attempting a cross-brand comparison is that most people won't have access to both platforms, so in its own way comparing Canon to Nikon is also an unfair one unless the photographer is buying into a platform for the first time... –  Jay Lance Photography Jun 3 '11 at 9:25
    
I'm more arguing from price-point then intention. There is a reason there are few full frame cameras intended for sports - They're slower. –  Fake Name Jun 4 '11 at 6:25
    
@Jay Lance Photography - Did I make any cross-brand comparisons? I don't think so. I am arguing from camera architecture, rather then specific camera implementations. I used canon camera as an example since I am more familiar with them, but what it comes down to is the amount of time it takes to shift out and digitize the number of electrons from the photo-sites, and all cameras work the same way in that regard (well, excepting Foveon sensors). Smaller sensors with fewer pixels can be read out faster, assuming the same ADC speed. –  Fake Name Jun 4 '11 at 6:28
2  
@Fake Name: Helloooo? Was I talking to you? My comment was clearly directed toward @eWolf (as is evidenced by the "@eWolf:" I put at the beginning of the comment)... He proposed a cross-brand comparison and it was to his comment that I was responding... –  Jay Lance Photography Jun 4 '11 at 8:25

In principal full frame cameras aren't bad for sports due to lack of reach. All of Canon's supertelephotos will take a teleconverter (or two) and there's always a longer lens you could get, until you hit 800mm with stacked teleconverters which is getting silly, even for sports.

You might have to pay more to get the same angle of view with full frame but full frame is more expensive in general, that applies to more than just sports.

It makes sense, but how high would the level of detail be on a full-frame with 300mm compared to the other example?

In both cases (same number of pixels, and same pixel pitch) you would expect the FF with a 300mm lens to give sharper images than a 1.5 crop with 200mm lens (assuming the lenses perform similarly). This is because in the same number of pixels case the larger FF pixels are less demanding of the lens's resolving power, and in the same pixel pitch case the FF simply has more pixels.

There's a more important factor than lens reach in this debate, however, and that is with a cropped sensor you get greater AF point coverage.

The AF chips of current DSLRs are based on FF models, meaning they take up more of the frame when you crop. This is a genuine advantage for sports, and one reason people have stuck with the oddball 1.3x crop Canon 1D series. You get a slightly larger viewfinder than APS-C making it easier to compose, and you get more of the frame covered with AF points, making it easier to track objects.

share|improve this answer
    
I don't see how sensor resolution is relevant in conjunction with lens resolution for determining the sharpness of the final image - given that the output format is the same? –  ysap Jun 3 '11 at 14:47
    
+1 for AF point density –  gjb Jun 3 '11 at 19:36

There's nothing inherently wrong about full-frame for sports photography, but I think what you're probably going to see is that cost factors make crop-sensor bodies a more competitive choice for semi-pro sports photographers.

There are some specific body lineup comparisons where you're going to see faster continuous shooting out of a crop body vs. a comparable full-frame body -- the Canon 7D vs. 5D-II is a great example of this. The same can (sometimes) be seen with features like autofocus -- again, the 7D lead the way vs. the 5D-II by introducing some autofocus advances before the 5D line. Both of these are are more a factor of which models are introduced in what order by the camera manufacturer than they are an indication of specific limitations in the full-frame format vs. cropped sensors (autofocus frame coverage notwithstanding).

I think it probably is fair, though, to assume that camera manufacturers understand that in those semi-pro lines, sports & wildlife shooters that gravitate toward the cropped-sensor bodies also really benefit from those features, so there's a greater demand to see those features first in the cropped-sensor bodies.

The biggest real factor favoring cropped-sensor bodies is the (approximately) 1.6 crop factor that makes a telephoto lens appear to be longer on a cropped-sensor body (although this isn't a true magnification, it emulates it in some ways -- see Matt's answer). This means that for a semi-pro sports photographer, they can grab a 70-200 f/2.8 lens for a 7D and get the same reach as they'd get with a 112-300 f/2.8 (if that existed) on a 5D. The lens equivalents don't really line up well in too many cases, but you can look at a 200mm f/2.8 prime in Canon's L line (msrp: $819) vs. a 300mm f/2.8 prime (msrp: $4879) to get a sense for the kind of difference this can make to a budget-sensitive photographer.

This factor breaks down a bit when you get to true pro photographers that can afford both the full-frame bodies and the glass to give them the reach they need. Scott Kelby does a fair bit of sports photography with a full-frame Nikon 3DS, and he's got some great blog posts that include discussions of equipment used:

Some of the shots in those collections are absolutely stunning, and make a pretty good argument for full-frame when you can afford the top-end equipment needed to make it work for sports photography.

share|improve this answer

My experience is that if you have a good camera the factors that will influence the quality of your sports/action photography are: 1) frames per second; and 2) higher ISO noise performance.

surfer image

This was shot on a 10FPS Canon 1DMkII. I just don't think I would have gotten the shot with my 1DsMkIII which only manages 3-5FPS. I don't think frame rate is a trick. I am good at predicting where a surfer will go on a wave and what he might do next, but I'll never know what the wave is going to do next. Most pro surf photographers are shooting crop sensors partially because of frame rate (and every pro news photographer I know is doing the same).

My advice to you is rent a couple of cameras, shoot some sports, and decide which you like best.

share|improve this answer

If you had two cameras identical in every way apart from the sensor size. There would be no difference in detail.

Think of the sensor as the film in a film camera - projecting the image onto 35 mm film compared with say 25 mm has no difference other than that the film is physically smaller and the image is cropped.

To answer the question title - Are full-frame cameras bad for sports photography? - I would say it depends on other factors - lenses available/weight of camera/money and a whole host of other variables

share|improve this answer

I mostly have two points to raise, one that doesn't seem to have been mentioned previously, the other more an extended comment on another answer.

My first point is that (at least with the bodies I've used) the full-frame and the crop-frame camera had almost the same pixel density on the sensor. That let me shoot at about the same, or just slightly longer focal length on the full-frame body while retaining about the same level of detail (or somewhat more) but getting a lot looser framing, so in fast-moving sports it was much easier to be sure of getting all the action in the frame. Then I could crop out the extra later. Sometimes I cropped about even all around, so it was essentially the same as if I'd been shooting with a cropped sensor.

Other times, however, I picked an off-center crop. In those cases, it's a pretty fair bet that if I'd been shooting with a cropped sensor, I just wouldn't have gotten the shot.

As far as speed of shooting (frame rate) goes, however, I have to disagree with @Fake Name. At first glance, frame rate seems like it would be important. I think a lot of beginners are "tricked" into spending more for cameras with high frame rates based (in large part) on "pro" cameras that also have higher frame rates.

At least in my experience, however, if you know what you're doing, frame rate is about 99% irrelevant. Consider an example:

enter image description here

This was at a professional baseball game, and that was a fastball moving close to 100 MPH. 100 MPH works out to ~147 feet per second, so even at 10 frames per second, your consecutive shots will have the ball nearly 15 feet apart. If you depend on a fast frame rate, you probably have to shoot practically every pitch for at least one entire game (and probably two or three) before you can depend on getting a shot like that.

That, however, was shot with a Konica-Minolta D7D, which had a maximum frame rate of about 3 FPS if memory serves -- and that was still completely irrelevant, because I had it in single-shot mode anyway. Despite (or really, because of) that I was able to capture pictures on this order about one out of three (or so) pitches where the batter actually swung so there was at least some chance.

Oh, as far as detail goes, that started as a 6 MP shot, and was then down-sized quite a bit. Realistically, its biggest shortcoming is that this was a night-game, and the D7D did not handle high ISOs well at all (so I was shooting at ISO 100).

I should add that this should not be read as my bragging about my incredible reaction speed or anything like that either -- quite the contrary, I'm pretty sure that most people can probably do about as well (at least with some practice) and there are undoubtedly at least a few who can do quite a bit better -- for that matter, I'm pretty sure I'd do better if I did it more often.

share|improve this answer

There is nothing inherently wrong with particular sensor sizes for sports photography; After all, in the pre-digital days, it was 35mm film in use.

Looking at Canon's camera aimed at sports photography, it has been optimised for speed of focus, frame rate as well as the build quality to cope with the knocks of professional use -- it still has a crop factor (using an APS-H sized sensor), but if there really were the benefits, we would most likely see a smaller sensor in there.

It would be difficult to do direct comparisons, as finding a full frame camera and a cropped sensor with the same number of pixels from the same manufacturer is hard enough, without bringing in other factors like advances in noise reduction and improved demosaicing algorithms in the camera itself (of course, you could use a zoom lens to get the same field of view) - only then could you categorically say whether the larger or smaller sensor performs better.

share|improve this answer

The paragraph you quoted assumes a full frame camera has the same amount of pixels as a small sensor camera. In this case the small sensor camera can be considered better for tele-photo, because a 200mm lens on it is equivalent to a 300mm lens on full frame. However the size of one pixel is also smaller in the small sensor camera. So this is a trade off between high ISO performance and tele-photo.

But if the two cameras have the same pixel density (or DPI) instead of the same amount of pixels, then you the small sensor camera is just a crop of the full frame, nothing else. Despite the price, in this case the full frame is always better.

share|improve this answer
    
@Matt Grum gives a few reasons for why it is more than just a simple crop of the FF, but otherwise your answer is fine. –  ysap Jun 3 '11 at 14:51
    
You mean the AF coverage? I never thought about it, and it's an interesting point. Small sensors may indeed have better AF coverage. –  Darksair Jun 3 '11 at 15:46

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.